Psychology Thesis Frequently Asked Questions (FAQ) Part 1
“I’m not sure what I want to do for my thesis.”
The issue is somewhat common because students do not think about what they want to do for their thesis. I often suggest them to think about a topic, even before Honours year. However, if they are already starting their thesis, it is still not too late, but they need to do some reading and at least get familiar with the field of study. Identifying the scope and topic of research is important because it is hard if you do not even know what you want to answer. Based on my experience, two major ways can help you to identify a topic of research.You can first read a bunch of articles, and then look for pattern and narrow your topic, which I call the bottom-up approach. Alternatively, you can first have a particular question in mind, and search more articles to expand from it, which I call top-down approach.1Students who adopt bottom-up approach are often followed by the top-down approach. If you are going to do the research for the whole Honours year, or even more for master or Ph.D., it is better to choose the topic you are interested in and capable of.
After you have a topic in mind, sometimes you find it has limited literature, and it leads to another issue:
“There is few/no past research about the topic.”
The common mistake students make is they often do a specific study, and search for literature with the same studies, but end up replicating the study without significant contribution (except for cultural difference, which is overdone in Malaysia).Moreover, I sometimes came across students who said they could not find the gap. When I asked why they did not do a particular idea, they said no study had done it, which can be a research gap because it is not supposed to be done, and that is how they can contribute. But they may ask:
“How do I support my hypotheses when there was no study?”
It is not the RESULT of the studies that support your hypothesis, but the RATIONALE of them. The literature you search do not have to be 100% match with your present study, and you can include any research that discuss the relationship you are studying, even they are not using the same variables. Remember, aliterature review is not only the summary of past research, but also the justification of your current study. For example, Facebook is a relatively new social media in the recent years, and you may want to explore the influence on stress level. Although not much research had directly done with it, many studies had investigated the relationship of blogging and diary-writing with stress, so you can use those to support your argument of Facebook having the relationship as well. To better support your argument, thetheoretical framework should be first identified to describe and explain the underlying mechanism of the relationship, and hopefully predict the result of your study, i.e. the hypotheses. The theoretical framework does not need to have a formal name, and it is just a statement or explanation of the relationship (even though based on my experience, you would get the name eventually in the process of research, but still, it is not necessary).With the same example, you believe Facebook may reduce stress because of social support in the study of blogging, or catharsis (emotional relieving) proposed by the research of diary-writing. Not only that, you can even do more than that by opposing the past studies, such as suggesting Facebook actually increases stress level because people would think their friends are happier according to theavailability heuristic.
Now you are collecting and analyzing the data in your study, and the common question I got was:
“My stats-test is …, what is my design?”
To answer the question, I don’t know, because it is your design determines the statistical test, NOT the other way around. The design is the general methodology of your research, and stats-test is only part of them. The two general designs in quantitative research are: Experiment when you manipulate a variable to establish causation, and non-experiment when only the relationship is studied without manipulation (Sometimes you may conduct a quasi-experiment when there is no randomization, equivalent control group, or non-manipulative IV).Regardless of the design, you have to specify your variables and how you study them to justify your stats-test. For instance, if you have two manipulative IVs with two levels each, and participants experience them independently with two DVs, then you have a 2 x 2 independent multivariate design. Sometimes the “name” of design has no agreement among researchers, let alone your lecturers, so it is still better to consult your supervisor. As I often tell the students, there is no “right” or “wrong” stats-test, only the appropriate one. Although you may want to use 2 x 2 independent MANOVA as what you learned, you can choose to do two independent ANOVAs, four t-tests for each IVs, or even twelve t-tests for all combination, if you can justify.
As you can see, you can have different statistical tests even with the same design, so I cannot tell you what your design is solely based on your stats-test, as it is possible you use ANOVA in experimental or non-experimental study. Wait… Did you say ANOVA in non-experimental design? Here is another question I often came across (and to be honest, it is the most annoying one, please don’t tell them this):
To be continued, click here for part 2
Featured header image: http://collegefun-d.com/
Jordan Oh (Veng Thang) is a 3rd year psychology student in HELP. He studied and has the experience in Education (Teaching Chinese as Second Language), and now is a member of Peer Mentors and PAL (Peer Assisted Learning) tutor in quantitative research and cognitive psychology. His interest is in soft science like statistics and psychology, especially about how people acquire knowledge and anxiety issue in academic setting, that’s why he loves the course. Also, he is gay.